What is Evidence-Based Medicine?

Importance of EBM in Clinical Practice

• Improves patient outcomes
• Enhances quality of care
• Promotes efficient use of resources
• Keeps practitioners updated with latest research

The EBM Process

The EBM Process: Ask

Formulate a clear clinical question - PICO

The EBM Process: Acquire

The EBM Process: Appraise

• Critically evaluate the evidence
• Assess validity, relevance, and applicability
• There are lots of tools for this:
  • Oxford Tools
  • JBI Tools
  • BMJ Tools

The EBM Process: Apply

The EBM Process: Assess

• Evaluate the outcome of the decision
• Reflect on the process and identify areas for improvement

Levels of Evidence

Question 1

Question 2

Question 3

Types of Clinical Research

Experimental vs. Observational Studies

• Experimental Studies
  • Interventions are actively manipulated.
  • Example: Randomized Controlled Trials (RCTs).
• Observational Studies
  • No intervention; researchers observe outcomes.
  • Example: Cohort studies, Case-control studies.

Study Designs

• Randomized Controlled Trials (RCTs)
  • Gold standard for assessing causality.
  • Random assignment to intervention or control.
• Cohort Studies
  • Observational; follows a group over time.
  • Can be prospective or retrospective.
• Case-Control Studies
  • Compares individuals with a condition (cases) to those without (controls).
  • Efficient for studying rare diseases.
• Cross-Sectional Studies
  • Snapshot at a single point in time.
  • Useful for prevalence studies.

Randomized Controlled Trials (RCTs)

Gold Standard for Causality

Cohort Studies

Observational Study Design

Case-Control Studies

Retrospective Study Design

Cross-Sectional Studies

Snapshot in Time

Critical Appraisal of Research Articles

Assessing Methodology

• Study Design: Was the study design appropriate for the research question? (e.g., RCT, cohort, case-control)
• Sample Size: Is the sample size sufficient to detect a meaningful difference or effect? (i.e. is there sufficient power)
• Blinding and Randomization: Were the participants and researchers adequately blinded, and was the randomization process robust?
• Outcome Measures: Are the outcome measures valid, reliable, and clinically relevant?

Critical Appraisal of Research Articles

%%{
  init: {
    'theme': 'base',
    'themeVariables': {
      'fontSize': '24px',
      'fontFamily': 'Arial, sans-serif',
      'lineColor': '#333',
      'primaryColor': '#f0f0f0',
      'primaryTextColor': '#333',
      'primaryBorderColor': '#666',
      'edgeLabelBackground': '#ffffff',
      'tertiaryColor': '#fff'
    }
  }
}%%
flowchart TD
    A[Did investigator assign exposures?] -->Y1(Yes)-->B[Experimental study]
    A -->N1(No)-->C[Observational study]

    B --> D{Random allocation?}
    D -->Y2(Yes)-->E[Randomised controlled trial]
    E --> E1["Regression / HR / T-test<br>ANOVA / Chi-Square / ITT"]
    D -->N2(No)-->F[Non-randomised<br>controlled trial]
    F --> F1["Regression / HR / T-test<br>ANOVA / Chi-Square / ITT"]

    C --> G{Comparison group?}
    G -->N3(No)-->I[Descriptive study]
    I --> I1[Summary statistics]
    G -->Y3(Yes)-->H[Analytical study]
    H ---> K[Cohort study]
    K --> K1["Regression / HR / T-test<br>ANOVA / Chi-Square / RR"]
    H ---> L[Case-control study]
    L --> L1["Regression (esp. Logistic) / T-test<br>Chi-Square / McNemar's / OR"]
    H ---> M[Cross-sectional study]
    M --> M1["Regression / T-test<br>Chi-Square / OR"]
    
    classDef yesno fill:none,stroke:none;
    classDef question fill:#cdc6ff;
    classDef tests fill:#ff9b9b;
    classDef study fill:#84cdee;
    
    class N1,Y1,N2,Y2,N3,Y3 yesno;
    class A,D,G question;
    class E1,F1,K1,L1,M1 tests;
    class E,F,K,L,M, study;

    %% Lengthen lines
    linkStyle 7,8,9 stroke-width:2px,fill:none,stroke:black;

Assessing Bias in Research

Common Types of Bias

• Selection Bias: Are the participants selected in a way that may skew the results?
  • Selection bias occurs when the participants included in a study are not representative of the target population, leading to results that may not be generalizable.
• Information Bias: Is there a systematic error in how data is collected or classified?
  • Information bias occurs when there is a systematic error in the way data is collected, measured, or recorded in a study, leading to inaccuracies in the data that affect the study’s outcomes.
• Confounding: Are there other variables that could influence the results, not accounted for?
  • Confounding occurs when an extraneous variable (a confounder) is associated with both the exposure and the outcome, leading to a spurious association between them.

Relevance to Clinical Practice

Assessing Relevance

Question 4

Question 5

Question 6

Role of Clinical Guidelines in Pediatrics

Introduction to Clinical Guidelines

• Definition: Clinical guidelines are systematically developed statements that assist clinicians and patients in making decisions about appropriate healthcare for specific clinical circumstances.
• Purpose: They aim to improve the quality of care by providing evidence-based recommendations.
• Structure: Virtually all guidelines begin with a systematic review of the literature.

Meta-Analyses (Part 1)

Overview and Key Concepts

• What is a Meta-Analysis?
  • Meta-analysis is a statistical technique that combines the results of multiple studies to produce a single estimate of the overall effect.
• Effect Size and Models:
  • Effect Size: A measure of the strength of the relationship between variables across studies (e.g., odds ratio, relative risk, mean difference).
  • Fixed-Effect Model: Assumes that all studies estimate the same effect size; variability is due to random error.
  • Random-Effects Model: Assumes that effect sizes vary across studies due to real differences; accounts for between-study variability.

Meta-Analyses (Part 2)

Advanced Techniques and Interpretation

• Forest Plots:
  • Visual representation of individual study results and the overall pooled estimate.
  • Each study is represented as a line, with the size of the marker indicating the weight of the study.
• Heterogeneity Assessment:
  • I² Statistic: Measures the proportion of variability in effect estimates due to heterogeneity rather than chance.
  • Cochran’s Q Test: Statistical test for heterogeneity; significant results indicate the presence of heterogeneity.
• Publication Bias Assessment:
  • Funnel Plots: Scatterplot of study effect sizes against sample size to detect asymmetry, which may indicate publication bias.

Interactive Funnel Plot

Interpretting Funnel Plots

:::

Interpretting Forest Plots

:::

Systematic Reviews (Part 1)

Overview and Process

• What is a Systematic Review?
  • A systematic review is a structured approach to reviewing the literature, aiming to answer a specific research question by synthesizing all relevant studies.
• Key Steps in Conducting a Systematic Review:
  • Formulate Research Question: Use PICO (Population, Intervention, Comparison, Outcome) framework to define the scope.
  • Search Strategy: Conduct comprehensive searches in multiple databases like PubMed, Cochrane, and Embase.
  • Study Selection: Apply inclusion and exclusion criteria to identify relevant studies.
  • Data Extraction: Systematically collect data from selected studies.

Systematic Reviews (Part 2)

Quality Assessment and Bias

• Critical Appraisal of Studies:
  • Risk of Bias: Evaluate potential biases in the included studies using tools like the Cochrane Risk of Bias tool.
  • Quality Assessment: Assess the quality of evidence using frameworks like GRADE (Grading of Recommendations Assessment, Development, and Evaluation).
• Synthesizing Evidence:
  • Narrative Synthesis: Summarize the findings from studies qualitatively.
  • Statistical Synthesis: If appropriate, use meta-analysis to quantitatively combine data.
• Common Biases in Systematic Reviews:
  • Publication Bias: Studies with positive results are more likely to be published.
  • Selection Bias: Inconsistent inclusion criteria across studies can affect outcomes.

Systematic Reviews (Part 3)

Example of a Bias Assessment by Cochrane